Wideband readings vs 5 gas readings.

[JTVOLVO]

I am concerned about using my wide band controller. When I set my AFR table to where I believe it is correct, use VE analyzer and run it on the chassis dyno, I find that the chassis dyno 5 gas analyzer shows that I am way too rich. The race engine likes about 13.1/1 AFR. This is where the engine makes the best torque curve and horsepower curve. Now remember I am mostly at full throttle and at 95-100 % MAP during these tests. I can calculate a correlation between the chassis dyno and my wideband and adjust my AFR settings and VE setting accordingly. I would rather that my wideband would correspond very close to the 5 gas readings. In my last race I set the wideband controller authority to 0 as I did not want it correcting my AFR during the race. I am not sure how to make both readings coincide so I can trust my settings when I change the settings at the track in Tuner Studio to my MS. I have read some about calibrating my wideband but am not clear about this. Any help is appreciated.

James

[devastator]

What kind of WBO2 are you using?

[JTVOLVO]

Zeitronix (Bosch).

[devastator]

I don't know how to calibrate this WBO2. Can you change the input voltage calibration settings in MT to give you the result you want? If this unit will output Lambda values, you could tune using Lambda instead of AFR.

[JTVOLVO]

Good idea Devastator,
I can do that but will have to look it up. If you know a quick link, let me know.

Thanks,

James

[devastator]

A link to change MT settings or tuning using Lambda?

[JTVOLVO]

Tuning using Lambda.

[devastator]

You have to go into the .ini files and set them up for Lambda values from you WBO2. Change MT to read Lambda values. Then you change your AFR table to Lambda values. It's pretty easy really. I did it a year ago, but, because I run alcohol, and MT calculates it's Lambda values, I had poor luck, and switched back to AFR. Then I programmed my Innovate WBO2 to output perfect Lambda as 14.7:1, and set everything else up accordingly, so my AFR table looks like it is for gasoline.

[JTVOLVO]

Ok, got it.
Thanks,

James

[FixItAgainTony]

James,
Running without the EGO on during a race assuming the tune is good is probably a good idea. As the tune gets better, the need for EGO correction should reduce. On any engine, once the tune is good, too much EGO authority will cause the engine to run too rich / lean in the event of an O2 sensor failure. I have seen this happen.

There are a few things going here. Let’s try to separate them.
1. Engine likes 13:1 based on the 5 gas analyzer.
2. O2 sensor does not agree with the 5 gas sensor (not too surprising)
3. Feedback to controller through for EGO

Let’s start with difference in the readings from the 5 gas sensor versus the O2 sensor. When trying to get accurate readings from sensors, it is necessary to calibrate them. In some cases, the only way to be absolutely sure that the readings are right is to run standards before and after the measurements to check if the system drifted. I would guess that both sensors [wide band O2 and the 5 gas analyzer] are wrong in the context just described, but I would also suspect that the 5 gas analyzer is much closer to the truth. This depends on when the standards were run, quality of the equipment and standards, how it was connected to the car, etc – assuming the person running the 5 gas analyzer is careful about these items, it will probably give the better result. Note that the result that the 5 gas analyzer gives max power 13:1 is consistent with what one would expect. In this instance, I would trust it more than the O2 sensor.

The max power measurement is an independent measurement from AFR. The absolute value of the AFR does not mean too much, as long as it is close [not melting pistons, etc] and one knows that at this AFR, measured this way gives the max power. As long as the AFR is measured in the same manner, the engine should produce the max power. ie, if the 5 gas sensor says 13.1:1, and it is actually (measured on golden million dollar equipment, 12.5:1) and 11:1 on your O2 sensor, the power will still be the same as long as the AFR for the given sensor used gives a consistent value
5 gas 13:1 = million$ 12.5:1 = O2 sensor 11:1 => all imply max power setting.

The O2 sensor likewise needs to be calibrated to standards to operate correctly. This is not something that is easily done… Most sensors can do a free air calibration and that is about it. They depend on a linear response (Y = mx +b; m = slope, b = Y-intercept) constant slope with the free air calibration determining the Y-intercept.
The wideband O2 sensors are also at the mercy of the environment they are in :
- electrical connections
- heat from the exhaust stream causes the internal sensor heater to operate in a range that is not optimal.
For a Bosch LSU4.2 sensor, the maximum temperature of the sensor at the bung (the sensor hexagon) should not exceed 500 C [ 900 F]. If these temperatures are exceeded a copper heat sink or extension should be installed … This is easily checked.
- clean air during operation to compare the exhaust stream.
From the Bosch datasheet:
9.13 For physical reasons the sensor needs ambient air at its reference gas side. Replacement of the air volume inside the sensor must be guaranteed by a sufficient air permeability of the wires and the connectors between sensor and ECU...
If the engine compartment is full of exhaust (not likely) the readings can be off.

Most likely the sensor is giving a value at a given load that is consistent, but off by a fixed amount (ie, O2 sensor reads 11.1, 5 gas analyzer reads 13.1). If this is the case, then a way to work around the problem is to change the equation for the O2 sensor so that the 5 gas sensor and the O2 sensor agree. Create the equation and then burn-it into the MS EGO table.

Another way to get there (perhaps easier) is to just enter in the O2 sensor AFR that corresponds to the 5 gas 13:1 value into the AFR table and be done with it, realizing that the O2 sensor has an offset in it. => it says 11:1 when the actual AFR is 13:1, so set the AFT table to 11:1.

When you adjust the Ve table to get the best readings the dyno, you need to be careful. If the gEGO correction was on, then you could be fighting your O2 sensor. If you were using the output of the 5 gas analyzer AFR into MS, then you should be fine, assuming the electrical connections were solid (ie, AFR read from MS = AFR read from the analyzer). This would require the calibration curve for the analyzer to be burned into the O2 sensor table, if it is different from your O2 sensor.

It would be useful to look at the data log from MS and the data log from the dyno w/5 gas analyzer for the same dyno pull. One could use a few points to build a new calibration curve for the O2 sensor that matches the 5 gas analyzer a little better.

If it were mine, I’d check the O2 sensor electrical connections / temperature during full load operation. If those were okay, then I’d just leave EGO turned off and use the car’s wideband to monitor what it thinks is the AFR. As long as it gives consistent readings at particular loads [not necessarily the correct reading, but reproducible, consistent readings] things should be okay.

-Charles.

[JTVOLVO]

Hi Charles and all,

Yes, it has taken quite a while for me to respond but I understand what you are describing. I will work on this in the off season as well as a good way to datalog while racing etc. I am now (at the end of this season) working on chassis enhancements for car handling as I have found more tweaking to be done there. I have one more race (and could mathmatically) win the championship. At the last race at Laguna Seca, my main competetor was there (the BMW) and we had a five lap shootout on the last five laps. He was fast on the slow tight corners and I was fast on the fast corners. He has a shorter wheelbase than I do hence the difference. It's good that the fast corners make up more time. He was on my bumper in turn 11, I slowed him down purposely on the entrance (knowing that he didn't have the torque I had coming out of the turn) and he moved to the inside for a pass, I gave him a small blocking move and dragraced him to the finish. I pulled away and won by a few car lengths (wonderful Volvo torque & MII v3!) I will keep you updated and still have one more chance to get that race lap record at Thunderhill this year.

James